In this study, we evaluated the effectiveness of the statistic, “Weighted Coherence” in relation to the average or mean coherence in a particular frequency band after cross-spectral analysis. In time domain, correlation coefficient explains the relationship between two variables of interest, and a similar statistic in frequency domain using cross-spectral analysis is referred to as coherence, which is equivalent to the squared correlation coefficient and ranges from zero to unity. In this study, we show that weighted coherence is much more sensitive to various challenges within the same subject and also may be a more sensitive indicator of group differences in different frequency bands. Our data on normal subjects and patients with panic disorder during supine and standing postures during normal breathing, and also at different breathing rates throw new light on the sensitivity of this previously described novel index to quantify coherence between two related time series such as beat-to-beat heart rate and QT interval.